Planning and configuration of self-contained energy systems for railways in China

Railway self-contained energy system (RSCES) is pivotal for sustainable rail development, yet existing studies lack a holistic methodology to systematically reduce grid dependence and fail to optimize configurations for dynamic grid-interaction capabilities. To bridge these gaps, this work first proposes a universal RSCES planning and configuration framework with adaptable topology and operation modes. Subsequently, it quantifies the critical impact of grid interaction constraints on configuration outcomes by employing an optimization method that combines an adaptive genetic algorithm with Yamlip/Gurobi. Finally, the configuration schemes for different operation modes are analyzed by combining with case studies of Hailesihao South Traction Substation, and the performance characteristics of the system configuration in various operating modes are also described. Simulation results indicate that Model A exhibits the lowest initial construction cost but higher long-term operational expenses, while Model B achieves lower life-cycle costs through electricity sales and residual value utilization despite higher upfront investment. Model C incurs the highest initial costs and self-contained rate but maintains moderate long-term operational expenditures. Opposite trends in photovoltaic configuration capacity and self-contained rates for Models A and B as grid interaction capacity decreases in grid-connected areas.